Usage¶

We start by initialising Wit using the token you’ll find under your Wit account Settings.

>>> from wit import Wit
>>> w = Wit('YOUR_TOKEN_WILL_GO_HERE')

To return an extracted meaning from a sentence, we use the get_message method...

>>> w.get_message('It puts the lotion in the basket')
{u'msg_body': u'It puts the lotion in the basket',
 u'msg_id': u'67771674-25ad-40af-b3fe-f2806c9a092a',
 u'outcome': {u'confidence': 0.525, u'entities': {}, u'intent': u'order'}}

...which is a simple wrapper around the GET /message API call.

You can send a sound file up for meaning extraction processing by passing a file-like object to the post_speech method. Like this:

>>> sound = open('hello_world.wav')
>>> w.post_speech(sound)
{u'msg_body': u'hello world',
 u'msg_id': u'6a410cda-32e0-4602-bcfb-c20f5e1aed66',
 u'outcome': {u'entities': {}, u'confidence': 0.525, u'intent': u'order'}}

Wit supports wav, mpeg3 and ulaw files. You’ll need to tell PyWit what the filetype is if it’s not a wav file.

>>> sound = open('hello_world.mp3')
>>> w.post_speech(sound, content_type='mpeg3')

See the POST /speech API docs for more info.

PyWit will output JSON data by default when there’s a response from Wit’s servers. If you want raw text, you can set that as follows:

>>> w.raw_text = True

You’ll find a couple more examples of cool stuff to try in the examples directory.

Most of the other Wit API calls have been implemented in PyWit, refer to the API Guide for examples.